public willingness to pay and policy preferences for tidal energy ... [PDF]

0 downloads 134 Views 123KB Size Report
source of bias in this study. ... 2 better understand how individuals value public goods and services, by ... renewable energy sources, including tidal energy.
           

PUBLIC  WILLINGNESS  TO  PAY  AND  POLICY  PREFERENCES  FOR  TIDAL  ENERGY   RESEARCH  AND  DEVELOPMENT:  A  STUDY  OF  HOUSEHOLDS  IN  WASHINGTON   STATE.       Hilary  J.  Polis1,  Stacia  J.  Dreyer,  and  Lekelia  D.  Jenkins   School  of  Marine  and  Environmental  Affairs,  University  of  Washington     Seattle,  WA,  USA      

1Corresponding  author:  [email protected]    

    INTRODUCTION       Puget   Sound   in   Washington   State   (WA)   has   significant  tidal  energy  resources,  but  the  industry   is  at  a  nascent  stage  of  development.  At  this  stage,   the   availability   of   research   and   development   (R&D)   funding   plays   a   critical   role   in   the   success   or  failure  of  renewable  energy  schemes.  However,   information   about   public   interest   in   developing   marine   renewable   energy   technology,   including   tidal   energy   technology,   in   WA   and   the   U.S.   has   been   limited.   Furthermore,   The   metrics   that   are   typically   used   to   value   marine   renewable   energy   projects   such   as   the   Levelized   Cost   of   Energy   (LCOE)  don’t  take  into  account  the  total  economic   value   and   non-­‐market   costs   and   benefits   of   developing  this  technology  [1].  A  recent  summit  of   ocean   energy   industry   stakeholders   identified   a   lack   of   quantification   of   the   total   economic   value   of   marine   renewable   energy   R&D   as   one   of   the   major   challenges   to   industry   development   [1].     The   objectives   of   this   study   are   two-­‐fold,   first   to   assess   public   preferences   for   potential   policy   incentives   and   funding   sources   to   support   tidal   energy   R&D   and   also   to   understand   the   non-­‐ market  values  associated  with  tidal  energy  R&D  in   WA   through   investigating   public   Willingness   to   Pay  (WTP).       METHODS     This   study   is   based   on   a   66-­‐question   mail   survey  that  was  sent  to  a  random  sample  of  3,000   WA   state   households.   We   employed   a   split-­‐sample   survey   technique   and   surveys   were   sent   out   to   1,500   coastal   and   1,500   non-­‐coastal   WA   state   households.   Coastal   residents   were   defined   as  

 

living  within  15  miles  of  the  Puget  Sound.  Marine   renewable   energy   technologies,   like   tidal   energy,   are   likely   to   impact   coastal   residents   and   non-­‐ coastal   residents   in   different   ways   and   we   wanted   to   understand   if   this   leads   to   a   difference   in   opinion  about  tidal  energy  development  (Petrova,   2010).   We   anticipated   that   non-­‐coastal   residents   would   be   under-­‐represented   in   our   sample   and   therefore   we   oversampled   non-­‐coastal   residents   to   ensure   that   we   could   make   comparisons   between   the   two   groups.     In   order   to   ensure   the   dataset   was   representative   of   residents   in   the   state  of  WA,  the  variable  of  coastal  residency  was   used   to   weight   the   dataset   according   to   pure   proportional   weighting   procedures.   We   sent   out   the   survey   in   early   July   2015   and   a   total   of   682   received   surveys   were   included   in   the   final   dataset,   which   yields   a   22.7%   response   rate.   Surveys   were   sent   out   with   a   cover   letter   describing   the   basics   of   tidal   energy   in   Washington.   We   recognize   that   an   overall   lack   of   public   knowledge   about   tidal   energy   could   be   a   source   of   bias   in   this   study.   A   non-­‐response   follow-­‐up   procedure   was   followed   and   respondents   did   not   have   significantly   different   levels   of   support   or   acceptance   for   tidal   energy   than   non-­‐respondents.   In   addition,   respondents   did   not   significantly   differ   from   respondents   for   significant  predictors  of  willingness  to  pay.       Contingent  Valuation  Methodology     Contingent  Valuation  Methodology  (CVM)  is  a   standard   non-­‐market   valuation   technique   for   renewable   energy   technology   [2].   CVM   combines   economic   theory   and   survey   methodology   to  

1  

better   understand   how   individuals   value   public   goods   and   services,   by   asking   them   how   much   they  would  be  WTP  [3].  The  CVM  question  on  this   survey   took   the   form   of   a   hypothetical   advisory   referendum.   Respondents   were   asked   if   they   would   be   willing   to   vote   “for”   or   “against”   a   measure   to   create   a   fund,   which   would   support   the   research   and   development   of   renewable   energy  technologies,  given  that  the  fee  for  creating   this  fund  would  increase  their  monthly  electricity   bills   by   a   certain   dollar   “bid”   amount   per   month.   Twelve   different   bid   amounts   ranging   from   $1-­‐ $100   loosely   based   on   Li,   Jenkins-­‐Smith,   Silva,   Berrens  and  Herron  [4]  were  rotated  through  the   surveys.   Bid   rotation   was   used   so   that   respondents   could   not   intentionally   bias   the   survey   by   writing   in   values   that   were   drastically   different   than   the   value   they   would   actually   be   WTP   [3].   In   order   to   understand   the   amount   that   respondents   would   specifically   be   WTP   for   tidal   energy   R&D,   participants   were   asked   a   follow-­‐up   question   about   what   percentage   of   the   fund   they   would   like   to   see   allocated   to   six   different   renewable   energy   sources,   including   tidal   energy.   Bid   amounts   and   responses   to   the   referendum   question  were  then  adjusted  to  be  specific  to  tidal   energy.       Addressing  Hypothetical  Bias     Hypothetical   bias   is   a   concern   when   respondents’   answers   to   a   survey   WTP   question   involving   a   hypothetical   scenario   are   different   than  they  would  actually  be  WTP  in  reality.  Champ   and   Bishop   [5]   explored   this   issue   by   comparing   answers   from   respondents   who   were   offered   an   opportunity   to   actually   pay   for   wind   energy   to   those   who   were   offered   a   hypothetical   opportunity,   and   found   the   amount   that   respondents  were  hypothetically  WTP  was  higher   than   what   they   were   actually   WTP.   They   concluded   that   asking   respondents   a   follow-­‐up   “certainty”   question   about   their   WTP   answer   and   then   recoding   “Yes”   votes   to   “No”   votes   for   respondents   who   indicated   that   they   were   uncertain   about   their   answers   could   reduce   this   hypothetical  bias.       Respondents   were   asked   a   follow-­‐up   question   regarding   their   level   of   certainty   about   their   response   to   the   WTP   question   on   a   scale   of   0-­‐10.   We   applied   the   uncertainty   recoding   methodology   from   Champ   and   Bishop   [5]   to   this   study,   in   line   with   previous   studies   on   WTP   for   renewable   energy  R&D  [4,  6].  In  keeping  with  the  procedure   used   in   each   of   these   studies,   “yes”   votes   were   recoded   to   “no”   votes   at   certainty   level   cutoffs   of   less  than  7,  less  than  8,  and    less  than  9  on  a  ten-­‐ point  scale.  To  illustrate,  if  data  was  recoded  at  the   8+   certainty   level   it   would   mean   that   if  

 

respondents   circled   a   7   or   below,   their   WTP   answer   would   be   recoded   as   a   No   and   if   they   circled   an   8,   9,   or   10,   their   WTP   response   would   remain   as   their   original   answer.   Maximum   Likelihood  Estimation  results  are  presented  using   the  raw  dataset  and  datasets  recoded  at  the  7+,  8+   and   9+   certainty   levels   for   comparison   (Table   3).     Champ  and  Bishop  [5]  found  that  recoding  results   at   the   8+   certainty   level   produced   the   best   estimates  of  actual  WTP.  Therefore  all  projections   in   this   study   were   completed   using   data   recoded   at  the  8+  certainty  level.     MAXIMUM  LIKELIHOOD  ESTIMATION     Maximum   Likelihood   Estimation   was   used   to   fit   an   exponential   probit   model   of   WTP   for   tidal   energy   R&D.   The   dependent   variable   was   the   “for/against”   response   to   the   WTP   question.   The   model   was   then   used   to   estimate   WTP   values   using   the   Krinsky   and   Robb   [7]   procedure   with   5,000  draws.           MODELING  WTP  FOR  TIDAL  ENERGY  R&D     The   results   discussed   in   this   section   are   reported   and   discussed   using   four   different   models.   The   first   model   includes   the   full   dataset   and   the   next   three   models   include   datasets   recoded   at   the   7+,   8+,   and   9+   certainty   levels   respectively   (Table   2).   Variables   such   as   knowledge   and   psychological   constructs   like   place-­‐attachment  were  tested  but  did  not  produce   the   best   model   fit   and   were   not   included   in   the   final  regression  output.       Previous   studies   have   looked   at   perceived   rewards  and  risks  of  wave  energy  development  in   Europe   at   the   level   of   an   individual   project   or   community   [8].   However,   no   previous   studies   have   measured   the   perceived   benefits   and   risks   of   tidal   energy   development   beyond   the   community   level  in  the  US.  Therefore,  risk  and  benefit  survey   questions   were   developed   from   lists   of   tidal   energy  researchers’  perceived  risks  and  benefits,  a   review   of   media   pieces   covering   various   stakeholder   groups’   opinions   on   the   proposed   tidal   project   in   Admiralty   Inlet,   and   feedback   on   perceptions   of   risks   and   benefits   from   a   focus   group   with   members   of   the   general   public.   Respondents   were   asked   about   their   level   of   agreement   on   a   series   of   statements   regarding   the   social,   environmental,   and   economic   risks   and   benefits   of   tidal   energy   R&D.     These   statements   were   used   to   create   indices   for   each   type   of   risk   and   benefit.   These   indices   were   standardized   for   comparison.   When   run   in   a   model   with   data   recoded   at   the   8+   certainty   level,   All   three   risk   indices   are   highly   significant   predictors   of   WTP   and   negative   and   all   three   benefit   indices   are   highly   significant   predictors   of   WTP   and   positive.  

2  

  The   variables   of   social,   environmental,   and   economic   risks   and   benefits   are   also   highly   correlated.  When  each  risk  and  benefit  index  was   run   in   a   separate   model   with   data   recoded   at   the   8+  certainty  level,  the  environmental  benefit  index   produced   the   best   overall   measures   of   model   fit.   In   order   to   address   concerns   about   multicollinearity,  the  environmental  benefit  index   was  the  only  risk  or  benefit  index  included  in  the   final  models  (Table  2).       Results   indicate   that   the   more   respondents   believe   that   tidal   energy   will   create   environmental,   economic,   and   social   benefits,   the   higher   their   probability   of   WTP.   Conversely,   respondents   with   stronger   beliefs   that   tidal   energy   R&D   will   create   economic,   social   and   environmental   risks   had   a   lower   probability   of   WTP.  Given  that  there  are  no  tidal  energy  devices   in   Puget   Sound,   there   is   a   lack   of   scientific   data   about   the   environmental,   economic,   and   social   impacts  of  tidal  energy  in  the  region.  It  is  hard  to   study  what  does  not  yet  exist.  The  results  suggest   that   in   the   absence   of   this   concrete   information,   participants’   WTP   to   invest   in   the   R&D   of   this   technology   was   heavily   influenced   by   their   perceptions  about  the  potential  risks  and  benefits   of  developing  tidal  energy.       Political  affiliation  and  income  were  generally   significant   predictors   of   WTP   across   all   certainty   models   (Table   2).   Therefore,   the   higher   the   respondents’   income   and   the   more   that   respondents   consider   themselves   to   be   liberal,   the   higher   the   probability   of   WTP.   The   coefficient   for   climate  change  index  was  significant  in  the  model   with   the   raw   dataset,   but   became   non-­‐significant   in   the   models   recoded   for   uncertainty.   Because   the   raw   dataset   may   suffer   from   hypothetical   bias,   we   can   not   conclude   that   the   belief   that   climate   change   is   a   human-­‐caused   problem   that   deserves   attention   contributes   to   a   higher   WTP   for   tidal   energy   R&D.   Previous   studies   have   shown   that   income  (Stigka  et  al.,  2014),  and  political  ideology   (Knapp   et   al.,   2013;   Wiser,   2007)   have   consistently   been   significant   predictors   of   WTP   for   renewable   energy   with   the   same   directional   relationships.   Additionally,   we   expected   coastal   residents   to   be   more   likely   to   benefit   from   tidal   energy   projects   and   thus   be   more   likely   to   WTP   for   tidal   energy   R&D   than   non-­‐coastal   residents.   The   variable   of   coastal   residency   was   not   significant   in   any   of   the   models,   indicating   that   this  is  not  the  case.       WILLINGNESS  TO  PAY  COMPARISONS   Results   indicate   that   the   median   amount   that   an   individual  respondent  would  be  willing  to  pay  for   tidal   energy   R&D   is   $1.62   per   month   (Table   3).   When  this  value  is  projected  to  reflect  the  amount  

 

that   all   2.9   million   households   in   Washington   would   be   WTP   it   equates   to   $57   million   annually   for  tidal  energy  R&D.       POLICY  PREFERENCES     Respondents   tend   to   prefer   that   the   federal   government   and   private   companies   should   be   most   responsible   for   funding   tidal   energy   R&D   (Table  1).       TABLE   1.   WHICH   ORGANIZATION   OR   INSTITUTION   DO   YOU   WA   RESIDENTS   THINK   SHOULD   BE   THE   MOST   RESPONSIBLE   FOR   FUNDING   TIDAL   ENERGY   RESEARCH  AND  DEVELOPMENT?    

Institution/Organization  

Percentage  

Federal  government  

37.5%  

Private  companies  

27.3%  

Public  Utility  District  

12.8%  

Other  

10.2%  

State  government  

9.1%  

None  

2.1%  

Local  government  

1.1%  

    Furthermore,   results   indicated   that   respondents   would   in   fact   like   to   see   the   federal   government   and   private   sector   work   together   to   develop   tidal   energy   technology.   When   respondents   were   asked   how   they   would   vote   on   a   series   of   hypothetical   ballot   initiatives   to   support  policies  that  have  been  used  to  fund  tidal   energy   development   in   other   countries   or   other   types   of   renewable   energy   in   the   United   States,   increasing   government   funding   for   Technology   Innovation   Systems   (TIS)   emerged   as   the   preferred  policy  support  approach  for  tidal  energy   development.   Nearly   78%   of   respondents   would   vote   “YES”   on   a   ballot   initiative   to   support   TIS.   TIS   have  shown  promise  for  marine  renewable  energy   development   in   Europe   and   could   help   support   marine   renewable   energy   through   the   ‘valley   of   death’   in   ways   that   other   market-­‐based   policies   cannot   through   the   creation   of   nursery   markets   and   acceleration   of   knowledge   diffusion   [9].   Respondents   were   less   likely   to   support   subsidy-­‐ based   policies   for   electricity   supplied   from   tidal   sources,   as   support   was   low   for   both   community   feed-­‐in-­‐tariff  and  contract  for  difference  policies.  A   green   loan   guarantee   program   was   similarly   unpopular.       A   second   WTP   follow-­‐up   question   was   asked   to   understand   respondents’   preferences   for   funding   different   renewable   energy   technologies.   Results   revealed   that   respondents   preferred   a   fairly   even   portfolio   allocation   of   R&D   funding   to  

3  

different   renewable   energy   technologies.     Respondents   believe   that   the   top   renewable   energy   technologies   to   receive   WA   state   R&D   funding   should   be   solar   (21%   of   funds),   tidal   (19%),   onshore   wind   (16%),   geothermal   (15%),   offshore   wind   (13%),   and   wave   (12%).     Further   research   should   be   done   to   understand   why   the   Washington   state   public   stated   a   preference   for   funding   tidal   energy   over   other   types   of   marine   renewable  energy  and  land-­‐based  wind  energy.     CONCLUSIONS     Recently,   in   conjunction   with   the   COP21   climate  talks  in  Paris,  many  private  investors  and   governments   pledged   unprecedented   support   for   renewable   energy   R&D.   This   is   likely   to   create   push   for   both   the   development   new   energy   technologies   but   also   demand   for   an   acceleration   of   bringing   these   technologies   to   market.   Results   from   this   study   demonstrate   that   for   tidal   energy   technology,  providing  R&D  funding  from  both  the   private   sector   and   federal   government   through   a   TIS  approach  would  likely  be  an  option  that  would   be  popular  with  the  public  in  WA.       This   study   also   found   that   the   previously   un-­‐ tested   attitudinal   constructs   of   perceptions   of   risks   and   benefits   are   strong   predictors   of   WTP   for  tidal  energy.  Interdisciplinary  collaboration  on   the   creation   of   these   indices   helped   to   capture   a   robust   picture   of   possible   risks   and   benefits.   Public  knowledge  about  tidal  energy  in  the  state  of   WA   is   relatively   low,   so   as   the   technology   advances   and   projects   start   going   into   the   water,   communicating  the  environmental,  economic,  and   social   risks   and   benefits   of   tidal   energy   development  to  the  public  will  be  important.       Results   show   that   the   WA   state   public   would   be   willing   to   pay   $57   million   annually   for   tidal   energy   R&D.   This   amount   is   almost   equivalent   to   the  annual  federal  budget  of  $60  million  allocated   to   marine   hydrokinetic   and   hydropower   R&D   in   fiscal   year   2015   through   the   Department   of   Energy   Waterpower   program   1 .     This   indicates   that   WA   state   residents   would   be   in   favor   of   a   significant   increase   in   tidal   energy   R&D   investment   over   current   public   spending   levels.   This   information   can   be   especially   helpful   for   the   tidal  energy  industry  as  it  provides  a  proxy  price-­‐ point  for  market  entry.       As  countries  and  private  donors  have  pledged   to   drastically   increase   the   supply   of   capital   available   for   clean   energy   R&D   in   the   years,   studies  like  the  analysis  presented  can  help  ensure                                                                                                                                           1  Obtained   from   the   U.S.   Department   of   Energy   Waterpower   Program   website   website:http://energy.gov/eere/water/water-­‐ power-­‐program-­‐budget    

that   funding   is   directed   in   a   way   that   aligns   with   societal   preferences   along   with   market   acceleration  objectives.       ACKNOWLEDGEMENTS     Funding   for   this   research   was   provided   by   a   Sustainable   Energy   Pathways   Grant   through   the   US   National   Science   Foundation,   NSF   Award   #1230426.   We   would   like   to   thank   Brian   Polagye   (University   of   Washington,   UW),   Terrie   Klinger   (UW),   Shannon   Davis   (The   Research   Group),   Katharine   Wellman   (Northern   Economics),   Robert   Berrens  (University  of  New  Mexico),  Julie  Mueller   (University   of   Northern   Arizona),   Abby   Towne   (ATowne   Design),   The   Sustainability   of   Tidal   Energy   Research   Team   at   UW,   Anita   Rocha   (UW   Center   for   Studies   in   Demography   and   Ecology,   CSDE),   Cori   Mar   (UW   CSDE),   and   Ted   Westling   (UW  Center  for  Statistics  and  the  Social  Sciences).     REFERENCES   [1]  Goldsmith,  J.,  2015,  "West  Coast  Regional   Strategies  for  Ocean  Energy  Advancement,"   Oregon  Wave  Energy  Trust.   [2]  Mitchell,  R.  C.,  and  Carson,  R.  T.,  1989,  Using   Surveys  to  Value  Public  Goods:  The  Contingent   Valuation  Method,  Resources  for  the  Future,   Washington,  D.C.   [3]  Carson,  R.  T.,  2000,  "Contingent  valuation:  A   user's  guide,"  Environmental  Science  and   Technology,  34,  pp.  1413-­‐1418.   [4]  Li,  H.,  Jenkins-­‐Smith,  H.  C.,  Silva,  C.  L.,  Berrens,   R.  P.,  and  Herron,  K.  G.,  2009,  "Public  support  for   reducing  US  reliance  on  fossil  fuels:  Investigating   household  willingness-­‐to-­‐pay  for  energy  research   and  development,"  Ecological  Economics,  68(3),   pp.  731-­‐742.   [5]  Champ,  P.  A.,  and  Bishop,  R.  C.,  2001,  "Donation   payment  mechanisms  and  contingent  valuation:   An  empirical  study  of  hypothetical  bias,"   Environmental  and  Resource  Economics,  19(4),   pp.  383-­‐402.   [6]  Mueller,  J.  M.,  2013,  "Estimating  Arizona   residents’  willingness  to  pay  to  invest  in  research   and  development  in  solar  energy,"  Energy  Policy,   53,  pp.  462-­‐476.   [7]  Krinsky,  I.,  and  Robb,  L.  A.,  1986,  "On   approximating  the  statistical  properties  of   elasticities  "  The  Review  of  Economics  and   Statistics,  68(4),  pp.  715-­‐719.   [8]  Bailey,  I.,  West,  J.,  and  Whitehead,  I.,  2011,  "Out   of  sight  but  not  out  of  mind?  Public  perceptions  of   wave  energy,"  Journal  of  Environmental  Policy  &   Planning,  13(2),  pp.  139-­‐157.   [9]  Corsatea,  T.  D.,  2014,  "Increasing  synergies   between  institutions  and  technology  developers:   Lessons  from  marine  energy,"  Energy  Policy,   74(0),  pp.  682-­‐696.  

4  

  TABLE  2.  WTP  FOR  TIDAL  ENERGY  R&D  MAXIMUM  LIKELIHOOD  ESTIMATION  RESULTS  

Variable   (Intercept)   AGE   EDUCATION   COASTAL   GENDER   CC_INDEX   Political   Affiliation   Income   BID  AMOUNT   ENVB_INDEX   Median  WTP   [95%  CI]   Mean  WTP   [95%  CI]   Pseudo-­‐R2   Log-­‐ Likelihood      

         

 

Full  dataset  

7+  Certainty  

0.400   (1.135)   0.009   (.008)   -­‐0.006   (0.110)   0.0209   (0.210)   -­‐0.279   (0.245)   0.264**   (0.135)   -­‐0.202**   (0.080)   0.0740   (0.058)   -­‐1.123***   (0.021)   0.782***   (0.138)   3.30   [2.69,  3.96]   4.90   [3.98,  6.48]   0.676  

0.368   (1.133)   0.006   (0.008)   -­‐0.139   (0.107)   0.119   (0.209)   -­‐0.202   (0.243)   0.152   (0.134)   -­‐0.199**   (0.079)   0.138**   (0.058)   -­‐1.056***   0.104   0.877***   (0.142)   2.20   [1.75,  2.70]   3.45   [2.77,  4.62]   0.643  

8+   Certainty   -­‐0.984   (0.163)   0.007   (0.008)   0.066   (0.107)   0.163   (0.207)   -­‐0.353   (0.246)   0.121   (0.132)   -­‐0.168**   (0.078)   0.161***   (0.059)   -­‐0.984***   (0.103)   0.867***   (0.142)   1.62   [1.23,  2.02]   2.71   [2.14,  3.72]   0.601  

-­‐94.588  

-­‐97.53  

-­‐100.388  

9+  Certainty   0.0648   (1.222)   -­‐0.005   (0.010)   0.061   (0.062)   0.164   (0.226)   -­‐0.265   (0.265)   0.094   (0.144)   -­‐0.134   (0.083)   0.037   (0.062)   -­‐0.953***   (0.112)   0.678***   (0.146)   0.96   [0.67,  1.27]   1.67   [1.30,  2.30]   0.563   -­‐81.851  

Notes:  1.  Standard  errors  are  in  parentheses  2.  *,  **,  ***  represent  significance  at  the  0.10,  0.05,  and  0.01   levels.  2.  Numbers  in  parentheses  are  standard  errors  3.  Bid  amounts  are  adjusted  to  the  80%  certainty     level  4.  Dependent  variable  is  the  binary  “for”  or  “against”  response  to  the  hypothetical  referendum     question  5.  95%  confidence  intervals  were  calculated  using  methods  from  [7]  6.  Values  of  mean  and     median  WTP  are  significantly  different  than  zero  for  all  models  7.  CC_Index  stands  for  Climate  Change   Index  and  EnvB_Index  stands  for  Environmental  Benefit  Index      

 

5